This paper reports some results of an Australian evaluation of six survey methods for collecting large-scale 24-hour travel data, and discusses a strategy to improve the quality of self-reported survey methods. A key aim of the first study was to design, implement and recommend a methodology, which gave high quality data, was cost-efficient and would be robust over five years. Five of the methods used a self-completion design, and had the advantage of low cost, but they had a low response rate and lower reporting of trips. In a survey, travel information needs to be collected on all trip stages by all travel modes. In the self-completion phase, there is an initial contact, a first mailing, and the posting of up to four reminders in the event of non-response. In the second study, self-completion was complemented by the use of telephone interviews, to clarify possible errors or uncertainties in responses, and interviews for a proportion of responding households, to check the quality and completeness of the self-completed data. A sample of non-responding households was selected for personal interviews to check reasons for non-response. A procedure for calculating non-reporting weights was developed and is described.
Abstract